Article Text

Download PDFPDF
Original article
Diagnostic accuracy of a two-stage model for detecting obstructive sleep apnoea in chronic tetraplegia
  1. Marnie Graco1,2,
  2. Rachel Schembri1,
  3. Susan Cross3,
  4. Chinnaya Thiyagarajan3,
  5. Shirin Shafazand4,
  6. Najib T Ayas5,
  7. Mark S Nash4,
  8. Viet H Vu5,6,
  9. Warren R Ruehland1,
  10. Ching Li Chai-Coetzer7,8,
  11. Peter Rochford1,
  12. Thomas Churchward1,
  13. Sally E Green9,
  14. David J Berlowitz1,10
  1. 1 Institute for Breathing and Sleep, Austin Health, Melbourne, Victoria, Australia
  2. 2 Department of Medicine, The University of Melbourne, Melbourne, Victoria, Australia
  3. 3 National Spinal Injuries Centre, Stoke-Mandeville Hospital, Aylesbury, UK
  4. 4 Miller School of Medicine, The University of Miami, Miami, Florida, USA
  5. 5 Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
  6. 6 Spinal Cord Injury Department, GF Strong Rehabilitation Centre, Vancouver, British Columbia, Canada
  7. 7 Adelaide Institute for Sleep Health: A Flinders Centre of Research Excellence, Flinders University, Adelaide, South Australia, Australia
  8. 8 Sleep Health Service, Southern Adelaide Local Health Network, Adelaide, South Australia, Australia
  9. 9 School of Public Health and Preventive Medicine, Monash University, Melbourne, Victoria, Australia
  10. 10 Department of Physiotherapy, The University of Melbourne, Melbourne, Victoria, Australia
  1. Correspondence to Ms. Marnie Graco, Institute for Breathing and Sleep, Austin Health, Heidelberg, VIC 3084, Australia; marnie.graco{at}


Background Obstructive sleep apnoea (OSA) is highly prevalent in people with spinal cord injury (SCI). Polysomnography (PSG) is the gold-standard diagnostic test for OSA, however PSG is expensive and frequently inaccessible, especially in SCI. A two-stage model, incorporating a questionnaire followed by oximetry, has been found to accurately detect moderate to severe OSA (MS-OSA) in a non-disabled primary care population. This study investigated the accuracy of the two-stage model in chronic tetraplegia using both the original model and a modified version for tetraplegia.

Methods An existing data set of 78 people with tetraplegia was used to modify the original two-stage model. Multivariable analysis identified significant risk factors for inclusion in a new tetraplegia-specific questionnaire. Receiver operating characteristic (ROC) curve analyses of the questionnaires and oximetry established thresholds for diagnosing MS-OSA. The accuracy of both models in diagnosing MS-OSA was prospectively evaluated in 100 participants with chronic tetraplegia across four international SCI units.

Results Injury completeness, sleepiness, self-reported snoring and apnoeas were included in the modified questionnaire, which was highly predictive of MS-OSA (ROC area under the curve 0.87 (95% CI 0.79 to 0.95)). The 3% oxygen desaturation index was also highly predictive (0.93 (0.87–0.98)). The two-stage model with modified questionnaire had a sensitivity and specificity of 83% (66–93) and 88% (75–94) in the development group, and 77% (65–87) and 81% (68–90) in the validation group. Similar results were demonstrated with the original model.

Conclusion Implementation of this simple alternative to full PSG could substantially increase the detection of OSA in patients with tetraplegia and improve access to treatments.

Trial registration number Results, ACTRN12615000896572 (The Australian and New Zealand Clinical Trials Registry) and pre-results, NCT02176928 (

  • sleep apnoea
  • clinical epidemiology
  • respiratory measurement
View Full Text

Statistics from


  • Contributors Study design: MG, RS, SS, NTA, MSN, WRR, CLCC, PR, TC, SG, DJB. Participant recruitment and data acquisition: MG, RS, SC, CT, SS, VV, WRR. Data analysis: MG, RS, WRR, TC, PR, DJB. Manuscript preparation and approval of final version: all authors. Guarantors: MG and DJB.

  • Funding MG was supported by an Australian Government National Health and Medical Research Council postgraduate scholarship (grant number 1114181) and an Australasian Spinal Cord Injury Network PhD scholarship. Data collection at Stoke-Mandeville Hospital was supported by a Stoke-Mandeville Masson Research Award grant from the Buckinghamshire NHS Trust Charitable Trust Funds, UK. Data collection at the University of Miami was funded by the Department of Defense (Award No W81XWH-13-1-0479).

  • Disclaimer None of the funders have had any role in the design, conduct, collection, management, analysis or interpretation of the data, nor have they been involved in the preparation, review or approval of the manuscript for publication.

  • Competing interests None declared.

  • Patient consent Obtained

  • Ethics approval Austin Health, Melbourne; GF Strong, Vancouver; NHS Health Research Authority, UK; The University of Miami, Miami.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles